AbdelStark / jev-benchmarks

git clone https://github.com/AbdelStark/jev-benchmarks.git

Probability-aware evaluation for typed decision models: calibration, selective risk, latency, and reproducible benchmarks.